1 Curle, J., A Roman frontier post and its people: the fort of Newstead in the parish of Melrose (Glasgow 1911)Google Scholar; available digitized at www.curlesnewstead.org.uk/

2 Id., “Roman drift in Caledonia,” JRS 22 (1932) 73-77.

3 Chapter 4 on Curle’s letters was published in limited circulation. Parts of chapts. 18 and 19 of the book under review appear in the Trimontium Trumpet, some volumes of which are available online at www.trimontium.org.uk.

4 Of the chapter headings in Curle (1911), only pottery vessels and the Newstead pits are not covered in some form by the new volume.

5 Other chapters too deal with the environs, particularly 9 and 17.

6 Ceramics from the fort are noted as requiring further study. The osteological and zoological remains still available would also benefit from updated study. Catalogues on iron, glass, and leather finds are forthcoming in the planned Newstead series, of which this volume is the first. One hopes that the final report on the Bradford University excavations will not be far behind.

7 Good locations to search for missing features (e.g., the bridge over the Tweed, cemeteries, temples, shrines, kilns) might become clearer once the final report from Bradford University, which conducted extensive geophysical survey around the fort, becomes available. On the other hand, some features may have been destroyed by the railway that cut through the site in the mid-19th c., as noted here by C. S. Sommer when he speculates (86) that the garrison’s baths might have lain in the S annexe, and by J. A. Smith in the 19th c., who suggested that the site’s ustrinum may have been destroyed during railway work: “Notices of various discoveries of Roman remains at Red Abbeystead, near the village of Newstead,” Archaeologia Scotica 4 (1857) 427 Google Scholar; whether Smith was correct, or simply misread a cremation cemetery for an ustrinum, the area he describes is near the course of Dere Street, which would be a logical spot for a cemetery.

8 In chapt. 7 on phasing, 8 on the annexes, 9 on the road network, 10 on the end of occupation at the fort, and 15 on military equipment.

9 National Museums of Scotland inv. no X.FV 29-30.

10 Gordon edited groups of Curle’s letters in ‘My dear Haverfield’: the Curle correspondence (1905-1909) (Melrose 2005)Google Scholar and ‘Letters to Hercules’: the Curle/British Museum correspondence (1891-1931) (Melrose 2008)Google Scholar [limited circulation publications].

11 The lack of a standard orientation, or at least a north arrow, in the figures (5.4-5.6) showing the new camp makes it difficult to read the photographs.

12 See the review of her work by Davies, J. L. in JRA 26 (2013) 760–63Google Scholar.

13 She also notes that the camps may have been not only for armies on the march, but perhaps could have been used by the garrison when rebuilding the fort (57), as happened at least twice.

14 See Clarke, S. and Wise, A., “Evidence for extramural settlement north of the Roman fort at Newstead (Trimontium), Roxburghshire,” ProcSocAntScot 129 (1999) 373–91Google Scholar.

15 His overall interpretation of the pits is as wells, since most would have reached the water table, although he also notes that cesspits have been found in similar numbers and some may have been used as latrines (80).

16 Even though figs. 9.2-9.3 use shading to differentiate known from suspected routes, it is not always clear from the discussion always clear from the discussion how certain the evidence is for the known routes versus how slender the evidence is for the suspected routes. Feature 7 varies noticeably between the two maps.

17 Clarke’s suggestion that the Antonine occupation may have begun sooner than previously thought due to a high level of Hadrianic coinage at the site is based on a faulty understanding of numismatics and coin circulation (100; see also chapt. 11).

18 Milne, A., A description of the parish of Melrose (Edinburgh 1743) 6 Google Scholar.

19 He includes 364 coins, 249 from the original publication, increased by later finds and fieldwalking, but lacking the “unpublished coins from the Bradford excavations” (123).

20 For his Currency Units, assignment of value to each coin follows the Roman system, where an as is worth 1, a sestertius is 4, a denarius is 16 and so on, tallying up the worth of the group of coins for each period being assessed and then comparing those sums. He leaves out aurei, which at a value of 400 each would skew the numbers too much, as well as copper alloys of indeterminate value.

21 This is something Curle was also interested in when he wrote about Roman objects on native sites.

22 See Keppie, L. J. F., Britannia 29 (1998) 381 nn. 94-95Google Scholar.

23 E.g., between Henig and Bishop; Hanson, Sommer, and S. Clarke; S. Clarke and Lonie; S. Clarke and Holmes; Holmes and Hunter.

24 Especially figs. 12.2-3, comparing coin distributions; fig. 16.4, no. 6; fig. 16.5, nos. 17-23; and fig. 16.6, illustrating enameled fibulae.

25 Excavations took place between 1989 and 1994, with further trenching until 1996, and geophysical survey up to 1998. The contrast between Curle’s excavation at Newstead, which produced interim reports as well as a much-lauded and still-appreciated final report the year after work ended, and Bradford’s excavation, which has failed to produce a final report in the two decades since the end of the main period of excavations, is conspicuous.

26 R. H. Jones, writing on the camps, received some information, but it is clear from what she reports that the Bradford excavations uncovered much more about the camps. Sommer relies extensively on brief reports produced by Bradford, while noting that not all were made available to him. Holmes does not include the coins they found. Hunter was able to include the 8 fibulae from the Bradford excavations in his table of fibulae by accessing an unpublished undergraduate thesis; this is remarkable considering that these finds must surely come to the National Museum of Scotland eventually, and if they had done so they would long since have been published.